New study finds our desire for ‘like-minded others’ is hard-wired

NEWS: A path-breaking new study on how we seek similarity in relationships, co-authored by researchers at Wellesley College and the University of Kansas, upends the idea that “opposites attract,” instead suggesting we’re drawn to people who are like-minded. The study could lead to a fundamental change in understanding relationship formation—and it sounds a warning for the idea that couples can change each other over time.
The investigation’s findings are presented in “Similarity in Relationships as Niche Construction: Choice, Stability, and Influence Within Dyads in a Free Choice Environment,” in the current issue of the Journal of Personality and Social Psychology, the field’s most respected journal. Assistant Professor of Psychology Angela Bahns (Wellesley College) and Professor of Psychology Chris Crandall (University of Kansas) are the paper’s lead authors.
In what might be considered a paradigm shift, the study’s most surprising discovery is that people in relationships change each other over time. Instead, Bahns and Crandall’s evidence places new emphasis on the earliest moments of a relationship—revealing that future friends or partners are already similar at the outset of their social connection, a major new finding, say the authors.
“Picture two strangers striking up a conversation on a plane, or a couple on a blind date,” says Bahns. “From the very first moments of awkward banter, how similar the two people are is immediately and powerfully playing a role in future interactions. Will they connect? Or walk away? Those early recognitions of similarity are really consequential in that decision.”
Whether or not a relationship develops could depend on the level of similarity the two individuals share from the beginning of their meeting. “You try to create a social world where you’re comfortable, where you succeed, where you have people you can trust and with whom you can cooperate to meet your goals,” Crandall said. “To create this, similarity is very useful, and people are attracted to it most of the time.”
Bahns adds, “Though the idea that partners influence each other is central in relationships research, we have identified a large domain in which friends show very little change— personality, attitudes and values, and a selection of socially-relevant behaviors.” She explains, “To be clear, we do not mean to suggest that social influence doesn’t happen in relationships; however, there’s little room for influence to occur when partners are similar at the outset of relationships.”
The data also suggests our drive to select like-minded others may be far stronger than previously assumed. “We’re arguing that selecting similar others as relationship partners is extremely common—so common and so widespread on so many dimensions that it could be described as a psychological default,” explains Bahns.
Bahns and Crandall stress the research shows people are not seeking shared similarity on one or two particular topics. “People are more similar than chance on almost everything we measure, and they are especially similar on the things that matter most to them personally,” Bahns said.
The study has major implications for how we grasp the foundations of relationships and approach relationships when the partners are different. Its findings were derived from real-world relationships. Data came from a field-research method dubbed “free-range dyad harvesting,” in which pairs of people interacting in public (romantic couples, friends, acquaintances) were asked questions about attitudes, values, prejudices, personality traits or behaviors that are important to them. The data were compared to see how similar or different the pairs were, and to test whether pairs who had known each other longer and whose relationships were closer and more intimate were more similar than newly formed pairs. They were not. Additionally, the researchers surveyed pairs who had just met (in a college classroom setting), and then surveyed the same pairs later. This allowed the benefit of longitudinal data, painting a picture of the same pairs over time.